Noscroll Wants to End Doomscrolling by Hiring an AI Bot to Do It for You

Add us on Google

There is a particular kind of exhaustion that has become one of the defining textures of being online in 2026. Not the exhaustion of ignorance, but of overexposure: the feeling that you have been on the internet for an hour and absorbed nothing useful while being thoroughly annoyed by everything you saw. Noscroll, a new startup that launched publicly this week, is betting there is a business in solving that problem by simply removing the human from the feed entirely.

The premise is straightforward: Noscroll is an AI-powered bot that browses your social feeds, news sites, and other online chatter, then texts you when something important happens. You do not go to the feed. The feed comes to you, stripped of everything the algorithm stuffed around it to keep your scrolling.

The company was built by Nadav Hollander, previously the CTO at the NFT marketplace OpenSea, who joined that company after it acquired his decentralized finance startup Dharma Labs in 2022. The leap from DeFi infrastructure to attention management tools is not as strange as it sounds. Both involve building systems that mediate between users and chaotic, incentive-distorted markets. In Hollander’s telling, the inspiration was personal. He designed Noscroll to address the overwhelming and often negative experience of being constantly online, where useful information is frequently mixed with distracting or stressful content. X, specifically, was the catalyst.

The product’s own pitch on the platform it was built to help users escape is telling: “X has the best information on the internet and the worst incentives and culture,” the company wrote. The goal is to surface the signal while cutting out the brainrot and ragebait. That tension, between a platform that is simultaneously the fastest way to learn something genuinely important, and a machine optimized to make you furious at strangers, is the exact gap Noscroll is trying to occupy.

Mechanically, the onboarding is deliberately low friction. To get started, you text the Noscroll AI agent directly at a phone number, and it sends a link to connect your X account, giving the service access to your likes, bookmarks, and the accounts you follow. The bot then pulls information from across the web beyond X, including news sites, blogs, Reddit, Hacker News, Substack, research papers, and more. You can also point it to specific sources you want monitored. It then texts your news digests at whatever frequency works for you a weekly roundup for casual readers, or multiple updates per day for those who need tighter coverage. Each digest includes links and a short summary, and you can reply to the bot to discuss what you’re reading or further refine your preferences.

That conversational loop is important. It is what separates Noscroll from the various RSS readers and newsletter aggregators that have tried to solve this problem before. Rather than a static list of sources that requires manual curation, the product is designed to learn from your responses and adjust over time. You tell it what mattered, what did not, and the digest sharpens accordingly.

The bot uses a variety of off-the-shelf AI models running on the company’s own proprietary infrastructure. The models have been customized with significant prompting, so the bot has its own distinct voice and communication style. Delivery happens via text message rather than an app or web dashboard, a choice that turns out to be the most interesting design decision in the whole product. Text messages feel bounded. You read them, they end. A feed, by design, never ends. Changing the delivery channel changes the psychological contract entirely.

Pricing is $9.99 per month, with a free sample digest and a seven-day trial that lets users customize preferences before committing. Subscriptions can be canceled at any time, and Hollander has noted that variable pricing may be introduced down the line.

The use cases emerging from the early user base are more eclectic than a pure tech-news product would suggest. Hollander says people are following niche anime industry news, local restaurant openings in Kyoto, job listings, layoff tracking, and more. That range is a signal that the core behavior Noscroll is monetizing, namely anxious, compulsive monitoring of a specific information stream, is not limited to the startup industry. Plenty of people have their own version of the feed they cannot stop checking and wish someone would summarize for them.

The competitive landscape here is genuinely sparse. There have been newsletter aggregators, AI-powered news briefings, and various attempts to build a calmer internet. None of them became culturally significant. The closest thing to a direct predecessor is probably the morning newsletter model popularized by The Hustle and Morning Brew, but those are editorially curated products aimed at broad audiences, not personalized monitors tuned to your specific mix of professional interests and guilty pleasures. Noscroll is closer to having a research assistant whose only job is to watch your feeds while you do something else.

The privacy question is worth sitting with, though. Connecting your X account and giving a startup access to your likes, bookmarks, and follows is a meaningful data handover. For a tool premised on removing the manipulation of algorithmic feeds, there is an irony in handing a new intermediary a detailed map of your online interests. Hollander has not yet addressed what happens to that data at scale or what the company’s data retention policies look like, and as Noscroll grows, that conversation will become harder to avoid.

The deeper question the product raises is whether solving doomscrolling at the personal level actually changes anything, or whether it just reroutes the same anxiety through a different pipe. The impulse to stay informed in a chaotic news environment is not irrational. What makes doomscrolling destructive is the infinite, unpredictable, reward-variable structure of the feed itself. Noscroll does change that structure in a meaningful way. Whether a text alert becomes its own compulsion, a different kind of checking behavior that carries its own anxious charge, is something only early users will discover over time.

For now, the product is a clean, well-framed solution to a real and widely-felt problem, built by someone with enough technical credibility to execute it and enough personal frustration to mean it.